As pretty much everyone who uses the Internet daily knows, Microsoft formally launched it’s revamped MSN Search site. This is a direct, intentional, focused, and admitted attack against everyone’s internet darling, Google (and probably Yahoo, but I can’t remember the last time I visited their site, so they’re irrelevant to me).

Personally, I think Wikipedia is doing something more interesting than MSN Search, but that’s because it’s different, innovative, extraordinarily useful, and basically a collective work of genius. Having more than one search engine is a good thing, sure, but it’s not interesting. But I digress…

I decided to do a side-by-side test drive of Google vs the New and Improved MSN Search. I compared straight up searches — number of results, first link returned, displayed Ads on the results pages. I got some unexpected results.

An Aside:

The first round of tests I did were a relevation, only in that I was on the UK MSN Search site and there was little indication that this was the case. I thought the results were a little off (what being UK-centric), but I didn’t realize the problem until I went to link to it and noticed the “.co.uk” URL bit. For some reason, I’ve never had this problem with Google, I’m just not entirely sure why.

While using the UK version of MSN Search first was an accident, it revealed something interesting — the results are different, even if you don’t use the “Only From United Kingdom” checkbox. What’s more is that some of the searches on the UK site produced more results than on the US site, which has no “Only From United States” checkbox. If the US version is the canonical “use this to search everywhere” site, why would it produce fewer results? If it’s not a canonical site, why doesn’t it have a “Only From United States” checkbox?

Microsoft, in response to people pointing out that they do not return as many results as Google has used the (very valid) defense that no one really has the time to dig through X million results. This is very true, and an interesting comment on the state of the Interweb (and the tools we use to find crap on it).

My results (summarized below), however, seem to indicate that this is more an “excuse” than an “innovation”, since some searches on MSN Search returned many more results than Google. Why the discrepency? Why on these particular search terms? Will MSN continue loading up its database until all terms return more results than Google?

Also, over the course of 30 searches, Google displayed 89 ads, most of which were relevant to the searches I was doing. MSN Search displayed 210, many of which were “correct” for the search term, but totally incorrect in reality.

For example, when I searched for “Thunderbird” on Google, there were no ads displayed, and the top link was to the Mozilla Thunderbird page. On the other hand, when I Googled for “Ford Thunderbird”, 3 car-related ads were displayed, and the top link was to the Ford site for their Thunderbird car. Searching MSN for “Thunderbird” did give me the Mozilla Thunderbird page as the first result, but it also displayed 11 ads about cars. “Ford Thunderbird” on MSN Search gave me the Ford site as top link, and the same 11 ads.

MSN needs to seriously work on their ad relevance. Also, redisplaying the ads from the top of the page at the bottom of the page? Totally unnecessary, and more than a little annoying. Stop sucking up so much of my browser’s real estate, thanks.

Ok, on to the summary of results. The following list contains the exact term I searched on in both Google (google.com) and MSN Search (search.msn.com), with the rough percentage of results MSN Search returned compared to Google. So if Google had 100 results, and MSN had 10, then the percentage is 10% (explanation for people like me who are really bad with numbers). Here’s the list:

• Linux – 43%
• Windows – 52%
• Mozilla – 33%
• .NET – 1.4% (??)
• Firefox – 5%
• Internet Explorer – 5%
• PowerMac – 50%
• OSX – 47%
• Apple – 53%
• iTunes – 19%
• Windows Media Player – 146%
• OpenOffice – 29%
• MS Office – 61%
• Thunderbird – 63%
• Outlook – 67%
• Outlook Live – 31%
• Donair Recipe – 10%
• Jeffrey Steingarten – 13%
• IBM – 44%
• SCO – 59%
• Internet Security – 178%
• Spyware – 23%
• Windows Vulnerabilities – 10%
• Linux Vulnerabilities – 13%
• Mac OS X Vulnerabilities – 10%
• Microsoft Security – 172%
• Open Source – 222%
• Free Software – 614%
• Google – 7%
• MSN Search – 63%

I don’t have any theories (conspiracy-like or no) about these results, I just find them odd as hell in some places. Presumably Microsoft is targeting certain areas of the web with their spiders, which would explain these results to some extent. I wonder how often their results will be updated and maintained. How are they sorting returned results — is it a formula akin to Google’s? Will people be able to cheat that formula through link spamming and such more easily than they can with Google?

I don’t know. Personally the ads on MSN are annoying enough to make me steer clear — during my experiment I accidentally clicked through on at least 5-6 ads on the MSN site, and none on Google. I also like my ads to be relevant and unobtrusive.

I also don’t particularly care for or trust Microsoft, so I’ll be sticking with Google, but you probably guessed that part already :)